Geek asks and answers

How to calculate the covariance matrix from each cluster, as from k-means?

I searched everywhere and I only found how to create a covariance matrix from one vector to another vector, for example cov (xi, xj). One thing that confuses me is how to get the covariance matrix from the cluster. Each cluster has many vectors. how to get them into one covariance matrix. Any suggestions?

information:

: vectors in the cluster, Xi = (x0, x1, ..., xt), x0 = {5 1 2 3 4} โ†’ the column vector

(in fact, this is a vector of MFCC functions, which has 12 coefficients per vector, after clustering them using k-means, 8 clusters, now I want to get the covariance matrix for each cluster to use as a covariance matrix in a Gaussian mixture model)

  output : covariance matrix n x n
+3
source share
1

: N D (, , "speaker1" ), D- point ( ", 1" ). : http://en.wikipedia.org/wiki/Multivariate_normal_distribution#Estimation_of_parameters http://en.wikipedia.org/wiki/Sample_mean_and_covariance

k = 8. , "" (, ), k = 8 . , k-mean ( , EM) , , , , "" , .

( , )

covariance , ,

a NxM, cov(x_i,y_j), X = (x1, x2,..., xN) Y = (y1, y2,..., Yn)

, , " ", ? - ... ? k=12 , ? ?

, , , x y E[x dot y] - (E[x] dot E[y]) ( ), , . .

, , . , , , .

+4

More articles:


All Articles